An empirical approach to insanity evaluations.

Described the development of the Rogers Criminal Responsibility Assessment Scales (RCRAS) as an empirically based testing procedure specifically designed for use in insanity evaluations. A pretest of the RCRAS that employed 10 case vignettes and a preliminary study of 25 patients are reported. Initial results indicated that the RCRAS has satisfactory interrater reliability and successfully discriminated between those patients evaluated as sane and insane. Results of the discriminant analysis, MANOVA, and factor analysis are discussed with reference to the RCRAS' construct validity and the need for further extensive studies.